As companies gather more data and seek to use it in meaningful ways, accessing that data and getting the most value from it can be a challenge.
In addition to the sheer volume of data, many of the decisionmakers and teams that need insights aren’t trained in analytics. To remain agile, a company’s business analysts, managers, executives, and other non-IT professionals need to leverage available data quickly, to answer new questions without the need for a lengthy IT development cycle.
Self-service analytics (SSA) offers a solution. By building an SSA program with a strong foundation of governed, tailored data and a visual analytics Center of Excellence (COE), organizations can democratize their data. SSA expands access by simplifying queries, even for users without technical skills, and by representing data visually so that it’s easy to understand at a glance.
SSA also helps solve the problem of data overload. A well-designed SSA program will have guardrails to surface only the relevant data for each question, so users don’t get overwhelmed or have to sort through tangential information while they look for answers. It also prevents the frustration that arises when queries take forever to run in the background while the executive or employee waits for answers.
Benefits of creating an SSA platform
SSA can deliver benefits to an organization on several levels. At the highest level, quick access to actionable insights helps the company to identify, address, and maximize their business opportunities as markets change. It gives decisionmakers the ability to reallocate resources, focus on new products and innovations, and pivot in other ways as conditions change.
On the individual contributor level, an SSA platform empowers employees to test their own hypotheses. They can access and explore their data in new ways and then present their findings with confidence that their data is correct. This capability helps individuals to add incremental value to the organization. It can also help them build their careers by asking and answering questions that add value.
For example, B2B salespeople can use their SSA platform to identify better ways to engage with and build offers for their customers. Forrester predicts that this year more than 60% of B2B sellers will be enabled by AI and automation.
The third area of benefits affects the IT department. By relieving the IT team from the work of database queries, SSA can free them to refocus on their core activities, like data preparation, integration, and optimization as well as infrastructure and security.
Identifying SSA user personas
To ensure that these guardrails are configured correctly, and to help each user get the most value from the platform, the organization’s SSA program must be defined and delivered in alignment with user personas. For example, executives need access to KPIs, trends, and exceptions as well as easy-to-use tools for filtering and sorting results, so they can make informed decisions quickly.
Managers need to adjust standard data dashboards so they can generate reports and answer questions as they arise. Analysts typically need to govern, tailor, and certify published data for specific use cases. They also often need to quickly create data visualizations and find insights fast.
All of these personas will want to ask their data questions in plain English. A good SSA will support this kind of Natural Language Query (NLQ) search, which is also called Natural Language Processing (NLP). In effect, it’s like having a Google search bar for your published and certified data sets.
Building the foundation for a successful SSA program
Four mission-critical elements must be in place to build an SSA program that works: data, data governance, training, and guardrails. The data from various sets has to be cleaned, integrated, and optimized. Data quality and relevance are more important than data volume, because the goal is to help users find accurate answers quickly.
Data governance and certification ensure that the data for the SSA program is reliable, accurate, and properly formatted. SSA data must be thoroughly validated so that the answers and insights it surfaces can guide decisions correctly.
Training on the SSA is important. Even though the visual format makes the data more accessible than traditional database queries, users need to know how to set search parameters at the very least. They may also need to know how to adjust dashboards on the fly or address new business questions, depending on their level of responsibility.
Finally, the tremendous power of SSA and visual analytics comes with a big responsibility: guardrails. Each user and group should have the level of permissions and access that are correct for their role. This helps keep users focused on the data that’s relevant to them. It also helps with database security.
Following the visual analytics maturity curve
With the foundational elements in place, organizations can move through the four stages of maturity with their SSA and visual analytics program. The first stage, reporting, is where many companies find themselves now. This stage features reactive data-use strategies, few data visualizations, static delivery of reports, and little or no automation.
Stage two, dashboards, adds traditional data visualizations like bar charts and trend lines, often presented on a dynamic platform, and some automation to refresh data. Sometimes we’ll see multisource data and some collaboration using visual data in stage two.
By stage three, we attain self-service. Employees are empowered to do their own ad hoc data research, data visualization options are more varied, and sharing and collaboration via dynamic platforms is more common. The data at this stage is certified, governed, and secure.
At stage four, AI enables predictive modeling and automated visualizations delivered by data scientists, developers, and analysts. As with the earlier stages, this stage relies on certified and secured data that’s automatically updated, but now users can search using NLQ. The system can also suggest next best actions and integrate with business processes.
Moving through all four stages can take as little as six to eight weeks but most organizations reach stage four within six months to a year after starting their program. Because of the time required to organize the data and move through each maturity level—and because of the need for businesses to work more efficiently, pivot faster and use data more effectively to make decisions fast—the time to consider developing an SSA is now.
About the author:
Derek Noce, Visual Analytics Portfolio Lead at Capgemini Americas
Derek Noce leverages decades of business analysis, strategic planning, analytics development, and deployment experience to deliver insightful visual analytics. He has led successful engagements across many industries, including travel services, digital advertising, financial services, healthcare, retail, consumer products & services, real estate, construction, and the public sector. Derek has a B.S. in Business Administration & Classical Studies from Boston University.